2°C: Beyond the limit

2°C: Beyond the limit

How we know
global warming is real

The answer includes Benjamin Franklin, Mutiny on the Bounty and centuries of records.

Sonnblick Observatory sits perched on a frigid summit in the Austrian Alps, 10,000 feet above sea level. It is one of the highest weather stations in Europe where meteorologists still live on the premises.

For all the ice and snow blanketing the observatory, it used to be much colder here. The station’s temperature records show a warming of 2.1 degrees Celsius (3.8 degrees Fahrenheit) since it opened in 1886, double the global average.

Back then, local gold miners trudged through waist-high snow drifts to carry construction materials up the mountain. The meteorological rock tower of the original observatory, fitted with key recording instruments, is still in use.

Meteorologists here have continuously recorded the temperature for 133 years — except for four days when troops left the station unmanned during World War I — and documented clear changes to climate. With millions of other measurements worldwide, their data help tell the story of global warming.

To reach the observatory, the crew hops on a cable car for a 10-minute ride up the mountain.

Sonnblick itself is now a victim of the very change it helps measure. Rising temperatures are melting the mountain’s permafrost, said Elke Ludewig, who heads the observatory. “And that’s why we had to put a lot of steel into the top peak, to save it from falling apart.”

Sonnblick Observatory opened and began recording the weather in 1886,  a technological wonder that measured how wind, barometric pressure and temperatures changed at upper altitudes, and during storms.

Click any temperature underlined in the story to convert between Celsius and Fahrenheit

Although the alpine station was advanced for its time, it was inspired by a compulsion as old as humanity itself: the need to monitor the weather, to understand it and to learn when to avoid it.

In the 19th century, violent storms sank so many ships on the Great Lakes that Congress authorized in 1870 a national weather service to help maritime navigators avoid peril. Eleven years earlier, a shipwreck off the coast of Wales led the United Kingdom’s newly formed Meteorological Service to set up a network of 15 coastal stations. Soon after, the forecasts generated were published daily in the Times of London.

“Much like what happens today, what prompts people to be interested in weather data?” said Jon Nese, a meteorology professor at Penn State. “Disasters.”

If early forecasting aimed to avert tragedy and economic loss, the troves of data it produced are used today to monitor a new sort of disaster, one that was scarcely foreseeable by 19th-century meteorologists but that now constitutes the single most significant fact about the planet’s environment. 

It is that the world is more than 1 degree Celsius (1.8 degrees Fahrenheit) hotter than it was before industrialization began pumping fossil fuels into the atmosphere. This warming has fueled new deadly fires, strengthened hurricanes and displaced people. And many areas have warmed far more than the average. 

How can that be known? How can it be possible to take Earth’s temperature, not just for this week or this year, but for decades and centuries? 

The answer begins with nearly 1,500 weather stations already operating by the time Sonnblick began recording. The telegraph allowed all those readings to be collected and analyzed to show weather patterns. 

In recent decades, meteorologists have relied on those records — and thousands more like them — to compare how the world’s climate used to be with how it is now. And as more observations from the past are retrieved from dusty archives worldwide, they point to the possibility of even more precipitous warming.

A worldwide web — of weather stations

Station density is key in a weather network. The closer the stations are to one another, the more comprehensive the weather records are.

2018
Distance from nearest
temperature observation

If Sonnblick were the only station on this map, meteorologists would not be able to accurately forecast the weather in places such as Paris or Berlin.

But Sonnblick is part of a dense network across western Europe, making it relatively easy to determine the continent’s temperature now, and over time.

These are all the stations on land that transmit data to the voluminous compilation of meteorological observations maintained by Berkeley Earth, a nonprofit research group.

The first land temperature observations in Berkeley Earth’s data set come from a Berlin station in 1701.

The first entry in North America is from Boston. That same year, the first weather reports in Russia come from St. Petersburg.

The other critical source of historical records comes from ships. Each of these points across the ocean shows where Benjamin Franklin recorded sea temperature as part of his study of the Gulf Stream, on his final Atlantic crossing, in 1785.

Starting in 1850, there are enough observations to produce accurate estimates of the planet’s average temperature for that year, and all subsequent years.

By 1900, Europe and the United States are covered in stations maintained by academics and amateurs, government workers and monks. Over the years, new clusters of stations follow the paths of colonization and trade routes — and the exploration of barely inhabited frontiers.

Five decades later, the network spans the world, although vast areas of South America and Africa still have far fewer stations.

Today, observations from more than 20,000 land-based stations, along with millions of readings from ships and buoys, feed into global temperature data sets.

At Sonnblick, where measurements are made every minute, the air gets so cold that Elke Ludewig and her crew must remove ice from frozen instruments. The thermometers transmit averages to Austria’s official Central Institution for Meteorology and Geodynamics, which routes them into key global temperature databases.

With that information, climatologists at independent and government agencies, such as NASA and the National Oceanic and Atmospheric Administration, map temperature patterns from the past to the present.

These maps rely on the fact that temperatures from nearby stations tend to correlate with one another over time — if the air around one station heats up, it probably also gets hotter at nearby stations. This correlation lets scientists pool the readings from multiple stations to calculate the temperature across the planet’s entire surface. In this way, scientists combine readings over space, and because the temperature record extends to the 19th century, they also are able to compare readings over time. 

There are complications. Older measurements were “taken in any number of weird or wonderful ways,” said Peter Thorne, a climate researcher at the National University of Ireland at Maynooth.

Some stations recorded temperatures in a variety of units now antiquated, such as Réaumurs, rather than Celsius; others fixed their prime meridian at Paris rather than Greenwich, England, the location of the now globally accepted prime meridian. Such inconsistencies force today’s climate scientists to double as historians, digging into the methodologies of old temperature records to determine the appropriate adjustments.

Past ocean temperatures can be particularly treacherous to calculate consistently. When sailors on merchant ships or explorer expeditions hoisted buckets from the ocean to record temperature, some of the saltwater would evaporate before reaching the deck. Because ships vary in size, the resulting difference in evaporation could produce changes in the temperature readings. Climatologists must adjust for these variations, too.

Some observations can’t be adjusted because they do not exist. During World War II, for instance, stations in Europe were forced to stop recording. Vast volumes of historical records have yet to be digitized; others have been lost or destroyed.

Despite these inconsistencies, and the fact that different research groups apply different adjustments, their conclusions show broad agreement on the degree of warming. There is such an abundance of reliable data now that any incremental adjustment or new method of computation can’t alter the overall number. 

“The bottom line is that any adjustments are relatively small in the global mean,” said Gavin Schmidt, who directs NASA’s Goddard Institute for Space Studies. “Neither are the calculated uncertainties anywhere close enough to affect the overall trends.”

[Extreme climate change has arrived in America]

The Washington Post analyzed four global temperature data sets to find the earth’s fastest-warming locales for our year-long series, 2ºC: Beyond the Limit. We found that roughly one-tenth of the globe has already warmed by more than 2 degrees Celsius (3.6 degrees Fahrenheit), when average temperatures over the past five years are compared with those of the mid- to late 1800s.  Catastrophic changes could be unleashed if that threshold is breached globally, a United Nations panel has warned.

Five-year average of temperature change compared with late 1800s

-1

0

1.5

2ºC

6+

Insufficient data

Extreme climate change is here, but it isn't distributed evenly. In Central Europe, Sonnblick Observatory lies in one of the fastest-warming areas of the world.

Off the coast of Uruguay, an ocean temperature spike has disrupted fisheries and nearly decimated populations of a seafood delicacy, the yellow clam.

The strong warming in the Gulf of St. Lawrence, amid the Canadian Maritime provinces, is rapidly melting the sea ice, and the Magdalen Islands there are crumbling into the ocean.

On Alaska’s North Slope, the town of Nuiqsut has warmed 4 degrees Celsius (7.3 degrees Fahrenheit) since the late 19th century. It is among the fastest-warming places on Earth.

Spin the globe to explore.

What if it’s even hotter?

What could alter the overall level of warming, however, and show that the planet has heated up even more than we now know, are records from more than a quarter-million ship logs and weather diaries stored in archives worldwide. 

From government initiatives to citizen science projects, thousands of people are working to recover and digitize these rapidly decaying records so they can be incorporated into existing data sets. Also imperiled are records throughout Africa, Asia and other former colonial outposts that are in danger of being destroyed. 

Some climate scientists theorize that they may discover additional warming when they incorporate these extensive records of weather between 1780 and 1850 or so, a period during which people already were burning some fossil fuels and clearing land of trees.  They suspect documenting old temperatures may reveal as much as a fifth of a degree Celsius (0.4 degrees Fahrenheit) of added warming.

“There are literally billions of observations which are still in paper format in various archives and libraries all over the world,” says Ed Hawkins, a climate researcher at the University of Reading in the United Kingdom.  

And half a million weather observations vanish every day, estimates Rick Crouthamel, who rescued data at NOAA and later formed the nonprofit International Environmental Data Rescue Organization (IEDRO) to continue the quest. “The paper is deteriorating, it’s rotting, it’s being eaten by rodents, the inks fade.”

The rescuers are keen to peer into enormous amounts of unretrieved and archived records in Africa, “proverbially a data desert,” as Thorne put it. And Hawkins is eager to transcribe logbooks of ships from the English East India Co., which plied trade on routes between Europe, Asia and India for decades in the 18th century. 

Almost all of this laborious deciphering needs to be done by hand, not by machine: The documents can be quirky and unpredictable, featuring sometimes complex notations and symbols.

Time

Date

Observations

Coordinates

Time

Date

Observations

Coordinates

Time

Date

Observations

Coordinates

Time

Date

Observations

Coordinates

Time

Date

Observations

Coordinates

Time

Date

Observations

Coordinates

Hour

Wind

Air temperature

Air temperature

Hour

Wind

Air temperature

Hour

Wind

Hour

Wind

Air temperature

Hour

Wind

Air temperature

Hour

Wind

Air temperature

The logbook kept by Capt. William Bligh of HMS Bounty shows how crews meticulously recorded meteorological observations in the late 1700s at sea. The numbers on these pages have not made their way into a database.

Although logbooks were not in a standard format, they typically had a date, a time, temperature, wind, barometric pressure — and longitude and latitude. So we know exactly where the observation was recorded in space and time.

This is the logbook page for April 21, 1789, which Bligh described as “very squally and dark gloomy” and “cloudy with a fresh gale.” One week later came mutiny on the Bounty, and the captain and his loyalists were set adrift at sea — but not before he grabbed his logbook.

Temperatures were recorded about every four hours, standard for Royal Navy logs. The temperature, noted in Fahrenheit, stayed consistently around 80 degrees.

The coordinates at the bottom of the page locate the ship near Tonga, a small archipelago about two-thirds of the way from Hawaii to New Zealand.

To rescue these records, researchers have used the sophisticated searchability of this era’s Internet to recruit volunteers passionate about the weather, as well as an earlier era’s efforts to measure and forecast it.

Crowd-sourcing projects such as Old Weather, Southern Weather Discovery and WeatherRescue rely on an elaborate human network of amateur historians, who go page by page to replicate the detailed and accurate observations of the past. Each annotation and interpretation is then double- and triple-checked by other volunteers and researchers. 

What has persisted over centuries, from seafarers to meteorologists, is a stringent requirement for precision that often reveals with dismaying clarity the disasters of weather and climate.  

Few places exemplify this as vividly as Sonnblick, where devotion to pristine record-keeping has been a source of pride for generations, and where the very mountain is melting beneath it.

Chris Mooney

Chris Mooney covers climate change, energy, and the environment. He has reported from the 2015 Paris climate negotiations, the Northwest Passage, and the Greenland ice sheet, among other locations, and has written four books about science, politics and climate change.

John Muyskens

John Muyskens is a graphics editor at the Washington Post specializing in data reporting.

Aaron Steckelberg

Aaron Steckelberg is a senior graphics editor who creates maps, charts and diagrams that provide greater depth and context to stories over a wide range of topics. He has worked at the Post since 2016.

Harry Stevens

Harry Stevens joined The Washington Post as a graphics reporter in 2019.

Monica Ulmanu

Monica Ulmanu is an assignment editor for the Washington Post. She joined The Post in 2018 from the Guardian newsroom in London.

Ann Gerhart

Ann Gerhart, senior editor at large, collaborates with journalists in video, photography and graphics to produce digital enterprise and to create new story forms.

Credits

Editing by Ann Gerhart.

Sources: Berkeley Earth; International Comprehensive Ocean-Atmosphere Data Set (ICOADS); The National Archives of the U.K. "The Sonnblick Mountain Observatory," Nature 49, 204–205 (1893). Ocean temperature observations from ICOADS shown on the map are from March only. The 3D model of the meteorological station was provided by the Sonnblick Observatory and adapted by the Post.

Share